Communication Protocol and Management Considerations for Internet of Things Gateways

Abstract

Today, it is inconceivable to think about modern life without the Internet. The Internet has transformed our lives, serving as a digital backbonefor communication, media and Web-based interaction among human users. It has, however, begun to evolve into the Internet of Things (IoT), interconnecting constrained devices such as sensors, actuators and miniaturised computing platforms. Aided in large part by advances in decreasing hardware costs, better battery, radio and computational technology, these kinds of connected IoT devices have rapidly begun to proliferate. The resulting traffic generated and exchanged by IoT devices is expected to dwarf that produced by humans.

However IoT devices exhibit significant differences in terms of computational and storage capacities, wireless radio networks and communication protocols. Integrating them into the Internet is not straightforward, in terms of device addressing, reachability as well as data representation and transfer. In many IoT domains, there are already deployed networks, devices and sensors based on legacy technology, that the IoT is compelled to integrate. All this has led to the emergence of IoT gateways in playing a central role for integration and interoperability, in ensuring interconnection and end-to-end communication with different kinds of IoT devices and networks.

IoT gateways themselves can be multi-functional; their complexity and usage depends on exactly what kinds of IoT devices and networks they need to integrate into the Internet, and what additional services are expected of them in ensuring successful interaction with edge devices. This dissertation focuses on these roles IoT gateways fulfill, enabling them to function as essential components of the IoT.

The main research question of this dissertation is: “For various kinds of IoT network topologies, how can gateways be configured and managed, to support connectivity and communication with IoT end devices?” To answer this question, four research areas for IoT gateways were studied, by applying Design Science Research Methodology (DSRM): Network Connectivity, Energy Consumption, Protocol Composition, and Gateway Management.

For network connectivity, the communication and reachability requirements exerted on IoT gateways by IoT devices and various IoT edge network topologies were studied. For energy consumption, the utilisation of gateway energy consumption patterns were investigated, in order to optimise communication as well as monitor operational performance. For protocol composition, the specification, implementation and deployment of protocols, gateway protocol stacks and network services was researched. Lastly, for gateway management, the research into redundancy management as well as development of IoT management practices for operational management and configuration, was performed.

The results and contributions of this dissertation are categorised into three abstraction levels of an IoT gateway. These are at the network, management and communication levels of abstractions respectively. At the network abstraction, the work is oriented towards IPv6based networks. The results describe the various IoT topologies, and gateway configuration and design to aid with IPv6 address allocation and network connectivity for edge devices. At the management abstraction, the dissertation identifies patterns in IoT gateway management and contributes object models and communication techniques for gateway configuration, operational monitoring, redundancy management and device proxying. At the communication abstraction, the dissertation contributes an implementation framework and a specification language allowing rapid development of communication protocols. The results also describe how energy consumption in IoT gateways can be employed both to optimise data transmissions as well as detect network-based attacks.

Both proofs of concept as well as conducted field experiments were used to verify the results of the dissertation. Empirical findings and obtained results have also been adopted both for Internet and IoT device management standardisation.